1. Field of the Invention
The invention generally relates to an image processing method, and more particularly, to a motion estimation method of dynamical images.
2. Description of Related Art
In an image processing course, every frame in a video image is divided into a plurality of macroblocks (MBs) and a whole compression program is broken up into a plurality of stages, in which each stage is in charge of processing a different compression program, for example, motion estimation and motion compensation program (MEMC), discrete cosine transform program (DCT program), variable length coding program (VLC program) and reconstruction program.
In a motion estimation course, the positions of the MBs of a motion object on a frame to be estimated are predicted according to the reference blocks of the motion object in the successive frames in association with newly-generated motion vectors, in which the motion amplitudes and motion directions of the MBs and the reference blocks are motion vectors.
In the prior art, during generating new motion vectors, the new motion vectors are often generated in stochastic refreshing way, which provides a random number sequence for output. The advantage of the above-mentioned scheme rests in easier implementation, but the disadvantage thereof rests in that in a same MB, different motion vectors are often unable to fast get their turns. In addition, when a motion object of an area in a frame instantaneously changes its motion path, the algorithm of motion estimation needs to fast converge; however, the conventional stochastic refreshing way for generation new motion vectors often fails to make the motion object fast converge to an accurate vector.
In this regard, it is necessary to provide a motion estimation method with faster convergence speed.